Redis: Replication, Part 3 – redis-py and Work with Redis Sentinel from Python

Still on the subject about Redis replication and Redis Sentinel – a couple of examples using the redis-py library for Python.

Previous series posts:

1. Redis: replication, part 1 – an overview. Replication vs Sharding. Sentinel vs Cluster. Redis topology
2. Redis: replication, part 2 – Master-Slave replication, and Redis Sentinel

All Redis clients for Python can be found here – redis.io/clients.

At first, we will use redis-py without Sentinel, just to check how it’s working.

Then – we will spin up Sentinel and check masters and slaves discovery.

A working environment aka Redis replication and Redis Sentinel are already configured in the previous post – here is just Python examples.

redis-py and Redis

Install redis-py on Debian:

root@redis-0:/home/admin# apt install python-redis

Also, pip can be used but I’d prefer to use normal repositories.

Check it:

root@redis-0:/home/admin# python
Python 2.7.13 (default, Sep 26 2018, 18:42:22)
[GCC 6.3.0 20170516] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> import redis
>>> r = redis.Redis(host='localhost', port=6379, db=0)
>>> r.get('test')
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
File "/usr/lib/python2.7/dist-packages/redis/client.py", line 880, in get
return self.execute_command('GET', name)
File "/usr/lib/python2.7/dist-packages/redis/client.py", line 573, in execute_command
return self.parse_response(connection, command_name, **options)
File "/usr/lib/python2.7/dist-packages/redis/client.py", line 585, in parse_response
response = connection.read_response()
File "/usr/lib/python2.7/dist-packages/redis/connection.py", line 582, in read_response
raise response
redis.exceptions.ResponseError: NOAUTH Authentication required.

Okay, works.

Find authorization in the documentation:

>>> help(redis.Redis)

Add the password:

>>> r = redis.Redis(host='localhost', password='foobared', port=6379, db=0)
>>> r.get('test')
'test'

Okay.

redis-py and Redis Sentinel

Now run Sentinel instances if they are not started yet and let’s try Redis via Redis Sentinel cluster client.

Sentinel’s configs /etc/redis/sentinel.conf on hosts now are:

sentinel monitor redis-test 52.58.69.184 6379 2
sentinel down-after-milliseconds redis-test 6001
sentinel failover-timeout redis-test 60000
bind 0.0.0.0
sentinel auth-pass redis-test foobared

After Sentinel instance start – it will update config, so now on the redis-1 (which is the first slave here), config looks like:

admin@redis-1:~$ cat /etc/redis/sentinel.conf
sentinel myid e3e62e6577aa975f93346dad3d4f8e25833fd8f1
sentinel monitor redis-test 52.29.101.118 6379 2
sentinel down-after-milliseconds redis-test 6001
bind 0.0.0.0
sentinel failover-timeout redis-test 60000
Generated by CONFIG REWRITE
port 26379
dir "/home/admin"
sentinel auth-pass redis-test foobared
sentinel config-epoch redis-test 1
sentinel leader-epoch redis-test 1
sentinel known-slave redis-test 52.58.69.184 6379
sentinel known-slave redis-test 35.159.18.26 6379
sentinel known-sentinel redis-test 172.31.46.202 26379 e0fe655c59aa3cc32eab1c0858c52418700abe79
sentinel known-sentinel redis-test 172.31.41.39 26379 07a450af0d2f178410b78ee0f5ae99ce1cd0ac62
sentinel current-epoch 1

Check Sentinel cluster status:

root@redis-0:/home/admin# redis-cli -p 26379 info sentinel
Sentinel
sentinel_masters:1
sentinel_tilt:0
sentinel_running_scripts:0
sentinel_scripts_queue_length:0
sentinel_simulate_failure_flags:0
master0:name=redis-test,status=ok,address=52.58.69.184:6379,slaves=2,sentinels=3

And use redis-py:

>>> from redis.sentinel import Sentinel
>>> sentinel = Sentinel([('localhost', 26379)], socket_timeout=0.1)

You can get a master’s IP using the discover_master() method:

>>> sentinel.discover_master('redis-test')
('52.58.69.184', 6379)

And slaves:

>>> sentinel.discover_slaves('redis-test')
[('52.29.101.118', 6379), ('35.159.18.26', 6379)]

To work with a master, use the master_for() method:

| master_for(self, service_name, redis_class=<class ‘redis.client.StrictRedis’>, 
connection_pool_class=<class ‘redis.sentinel.SentinelConnectionPool’>, **kwargs)
| Returns a redis client instance for the "service_name" master.

>>> master = sentinel.master_for('redis-test', socket_timeout=0.1)

But if call master now – it will tell us we aren’t authorized:

>>> master.set('test-key', 'test-value')
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
File "/usr/lib/python2.7/dist-packages/redis/client.py", line 1072, in set
return self.execute_command('SET', *pieces)
File "/usr/lib/python2.7/dist-packages/redis/client.py", line 573, in execute_command
return self.parse_response(connection, command_name, **options)
File "/usr/lib/python2.7/dist-packages/redis/client.py", line 585, in parse_response
response = connection.read_response()
File "/usr/lib/python2.7/dist-packages/redis/sentinel.py", line 55, in read_response
return super(SentinelManagedConnection, self).read_response()
File "/usr/lib/python2.7/dist-packages/redis/connection.py", line 582, in read_response
raise response
redis.exceptions.ResponseError: NOAUTH Authentication required.

So while creating the master object – add password:

>>> master = sentinel.master_for('redis-test', socket_timeout=0.1, password='foobared')

Check now:

>>> master.set('test-key', 'test-value')
True

Also, you can get a whole master’s configuration:

>>> master.config_get()
{'appendonly': 'no', ... 'slave-announce-port': '0'}

Or in a more readable view:

>>> for i in master.config_get():
...   print(i)
...
appendonly
requirepass
daemonize
protected-mode
zset-max-ziplist-entries
zset-max-ziplist-value
dir
slave-serve-stale-data
cluster-require-full-coverage
slowlog-log-slower-than
masterauth
rdbchecksum
...

On a salve – check the key we added previously:

admin@redis-1:~$ redis-cli -a foobared get test-key
"test-value"

Work with slaves is similar to master:

>>> slave = sentinel.slave_for('redis-test', socket_timeout=0.1, password='foobared')
>>> slave.get('test-key')
'test-value'

Master Change and Sentinel Failover

Let’s check out the master at this moment:

>>> sentinel.discover_master('redis-test')
('52.58.69.184', 6379)

And eventually, let’s try to stop Redis master on the redis-0 host:

root@redis-0:/home/admin# systemctl stop redis-server

Or by:

root@redis-0:/home/admin# redis-cli -a foobared DEBUG sleep 600

(if you’ll just kill the redis-server process with the kill – Sentinel will restart it again).

Check Sentinel’s logs:

10633:X 01 Apr 13:46:10.430 # +sdown master redis-test 52.58.69.184 6379
10633:X 01 Apr 13:46:10.486 # +odown master redis-test 52.58.69.184 6379 #quorum 2/2
10633:X 01 Apr 13:46:10.486 # +new-epoch 1
10633:X 01 Apr 13:46:10.486 # +try-failover master redis-test 52.58.69.184 6379
10633:X 01 Apr 13:46:10.488 # +vote-for-leader e3e62e6577aa975f93346dad3d4f8e25833fd8f1 1
10633:X 01 Apr 13:46:10.492 # 07a450af0d2f178410b78ee0f5ae99ce1cd0ac62 voted for e3e62e6577aa975f93346dad3d4f8e25833fd8f1 1
10633:X 01 Apr 13:46:10.492 # e0fe655c59aa3cc32eab1c0858c52418700abe79 voted for e3e62e6577aa975f93346dad3d4f8e25833fd8f1 1
10633:X 01 Apr 13:46:10.559 # +elected-leader master redis-test 52.58.69.184 6379
10633:X 01 Apr 13:46:10.559 # +failover-state-select-slave master redis-test 52.58.69.184 6379
10633:X 01 Apr 13:46:10.611 # +selected-slave slave 52.29.101.118:6379 52.29.101.118 6379 @ redis-test 52.58.69.184 6379
10633:X 01 Apr 13:46:10.612 * +failover-state-send-slaveof-noone slave 52.29.101.118:6379 52.29.101.118 6379 @ redis-test 52.58.69.184 6379
10633:X 01 Apr 13:46:10.664 * +failover-state-wait-promotion slave 52.29.101.118:6379 52.29.101.118 6379 @ redis-test 52.58.69.184 6379
10633:X 01 Apr 13:46:11.498 # +promoted-slave slave 52.29.101.118:6379 52.29.101.118 6379 @ redis-test 52.58.69.184 6379
10633:X 01 Apr 13:46:11.498 # +failover-state-reconf-slaves master redis-test 52.58.69.184 6379
10633:X 01 Apr 13:46:11.557 * +slave-reconf-sent slave 35.159.18.26:6379 35.159.18.26 6379 @ redis-test 52.58.69.184 6379
10633:X 01 Apr 13:46:12.550 * +slave-reconf-inprog slave 35.159.18.26:6379 35.159.18.26 6379 @ redis-test 52.58.69.184 6379
10633:X 01 Apr 13:46:12.550 * +slave-reconf-done slave 35.159.18.26:6379 35.159.18.26 6379 @ redis-test 52.58.69.184 6379
10633:X 01 Apr 13:46:12.632 # -odown master redis-test 52.58.69.184 6379
10633:X 01 Apr 13:46:12.632 # +failover-end master redis-test 52.58.69.184 6379
10633:X 01 Apr 13:46:12.632 # +switch-master redis-test 52.58.69.184 6379 52.29.101.118 6379
10633:X 01 Apr 13:46:12.632 * +slave slave 35.159.18.26:6379 35.159.18.26 6379 @ redis-test 52.29.101.118 6379
10633:X 01 Apr 13:46:12.632 * +slave slave 52.58.69.184:6379 52.58.69.184 6379 @ redis-test 52.29.101.118 6379
10633:X 01 Apr 13:46:18.647 # +sdown slave 52.58.69.184:6379 52.58.69.184 6379 @ redis-test 52.29.101.118 6379

Check from Python:

>>> sentinel.discover_master('redis-test')
('52.29.101.118', 6379)

Master was changed – 52.29.101.118 – it’s the redis-1 host.

Check status from the redis-1:

admin@redis-1:~$ redis-cli -a foobared info replication

Replication role:master connected_slaves:1 slave0:ip=35.159.18.26,port=6379,state=online,offset=76847,lag=0 ...

Check the Redis node status on the redis-0 host – it must become a slave now:

root@redis-0:/home/admin# redis-cli -a foobared info replication
Replication
role:slave
...

And try to add a new key from Python on the redis-0 host using the same master object we created before:

>>> master.set('test-key2', 'test-value2')
True

Check:

admin@redis-2:~$ redis-cli -a foobared get test-key2
"test-value2"

Done.